A fusion and prediction method of landslide multi-field monitoring data based on big data thinking

Abstract

The invention discloses a method for fusion and prediction of multi-field monitoring data of landslides based on big data thinking. According to the method, various obtained monitoring data is the unique basis, and a landslide slip prediction function implied in various monitoring data is determined through integration variance calculation, correlation analysis, cluster analysis, regression analysis and BP neural network analysis; multivariate statistical quantitative analysis and qualitative trend analysis of a BP neural network are combined and mutually complementary and verified, data categories participating the prediction function and a preliminary prediction function are determined through multivariate statistical analysis, and the reasonability of the prediction functions is verified through trend simulation of the BP neural network, so as to adjust participant data categories, make the data categories unified ideally and finish slip prediction of each landslide. The technology can be used for real-time prediction of the landslides, nationwide landslides are subjected to qualitative census and weather forecast early warning, and the method has great practical significance and wide application prospects.

Description

technical field [0001] The invention relates to disaster prediction and early warning technology, in particular to a fusion of landslide multi-field monitoring data and a slip prediction method based on big data thinking. Background technique [0002] The existing landslide slip prediction generally adopts the following methods: [0003] 1. Multivariate statistical analysis: It is a branch of mathematical statistics, which mainly studies the statistical regularity of interdependence between multiple variables (or multiple factors) in objective things. Important multivariate statistical analysis methods include: regression analysis, discriminant analysis, cluster analysis, principal component analysis, correspondence analysis, factor analysis, correlation analysis, variance analysis, etc. For example, correlation analysis can analyze the closeness between variables, cluster analysis can classify the degree of influence of data on the basis of similar analysis, and regression...

AI Technical Summary

Problems solved by technology

The learning and memory of the network is unstable, which has a lot to do with the type of samples and the number of samples learned. The selection of effective sample types is crucial to whether the output results are reasonable, and this is beyond the power of the neural network itself. In addition, BP The neural network only provides trend prediction, and generally there is no clear prediction function

Images